Our long-term goal is to define the neural substrate that is responsible for the central generation and control of voluntary movement. In this proposal we will explore one major component of this substrate- the primary motor cortex (M1). In the course of recent anatomical experiments, we discovered that M1 is not a uniform area, but rather has two subdivisions. The New subdivision of M1 is only present in some monkeys, great apes and humans. It contains output neurons with axons that descend to the spinal cord and make direct connections with motoneurons. These neurons are termed cortico-motoneuronal (CM) cells. In contrast, the Old subdivision of M1 is the standard for most mammalian species. Old M1 contains output neurons that descend to the spinal cord, but influence motoneurons only indirectly through interactions with spinal interneurons. The demonstration that M1 contains two subdivisions raises an important question- Do the aspects of movement represented and controlled in Old and New M1 differ? Physiological methods exist for identifying CM cells. In addition, we have a unique behavioral paradigm which enables us to define the aspect of movement encoded in the activity of cortical neurons during wrist movement. Our paradigm, along with the physiological identification of CM cells, will allow us to answer a critical question- What is the nature of the descending commands that CM cells send to motoneurons? To address these questions, we will train monkeys on our paradigm which dissociates 3 reference frames associated with movement generation- muscle activity, joint movement (intrinsic parameters) and direction of action (an extrinsic parameter). Then, we will record the activity of single neurons in M1 during the task. We will sample Old M1 and New M1 which lie in cortex on the surface of the precentral gyrus and buried in the anterior bank of the precentral sulcus. In addition, we will record the activity of 10-12 chronically implanted muscles in the forearm. We will use spike-triggered averaging of the muscle activity to identify CM cells. The results from this analysis will enable us to determine: i) the nature of the motor commands issued by M1 neurons with a direct influence on motor behavior and ii) the aspects of movement represented and controlled in Old and New M1. Overall, the results from our proposed experiments will provide novel information about the functions of M1, its subdivisions and its role in the planning, generation and control of movement.